AMOUNT OF PHOSPHORIC ACID IN THE SEA-WATER. 125 
times the weight of the phosphorus pentoxide, while iron should not 
exceed 1200 times its weight. They recommend also that for the greatest 
accuracy two standard solutions should be made up, one containing 
0:03 mg. P,O,; for use with samples containing this amount or more, 
and another containing 0-02 mg. for samples containing less than 0-03 mg. 
Colorimeter. The writer has used the Dubosq pattern with comparison 
tubes 5 em. in height and a swinging shade in front to cut off side light. 
For the comparison of phosphorus samples the model with 10 em. tubes 
would probably have been better. 
Filter papers should be washed with dilute nitric acid and hot water. 
The writer has found traces of phosphoric acid in two of the best-known 
hydrofluoric-acid washed papers. 
Porcelain and glass should be tested before use by extraction with hot 
dilute nitric acid and dilute ammonia. There are some varieties which 
will give up several milligrams of P,O; during an analysis. 
India-rubber should not be allowed to come in contact with the acid 
or alkaline solutions. 
The analyses have been carried out as follows: The samples were 
taken in glass-stoppered “ Winchester quart” bottles, holding about 
2700 com. As they were collected so near to the Laboratory it was 
generally possible to begin the filtration within three hours, so that there 
was no necessity for sterilisation. Filtration was carried out by replacing 
the glass stopper by one of rubber through which passed two glass tubes, 
which projected about 6 mm. on the inner side ; outside the bottle one 
projected about 25 or 30 mm., the other a few millimetres less. The 
bottle, full to the stopper, was quickly inverted on a retort ring with the 
tubes projecting into the filtermg funnel below the upper edge of the 
filter paper. Filtration then goes on without attention ; as a rule it was 
started in the afternoon and was finished by the following morning. 
Double papers were always used, sometimes Schleicher and Schull, No. 
589, “‘ black band” inside, to catch the coarser particles, with a “ blue 
band” outside ; at other times Whatman’s papers, No. 1 or No. 40. If 
there is much sediment No. 40 is almost too slow; one sample took 
thirty-six hours to filter. 
As a rule 500 cem. was taken. The water was measured into a hard 
glass beaker. and 10 cem. of 2 N ammonium chloride and 1 cem. of the 
iron mixture were added, with a few drops of dilute nitric acid to dissolve 
the precipitate. The mixture was heated to 70° or 80°C. on the water- 
bath and precipitated with the smallest possible quantity of dilute am- 
monia ; the heating was continued until the precipitate had collected 
